Removing noise from cardiac signals

Abstract

In one embodiment, a method includes receiving first cardiac signals captured by at least one first sensing electrode in contact with tissue of a first living subject, injecting the received first cardiac signals into a length of wire, which outputs respective noise-added cardiac signals responsively to noise acquired in the wire, training an artificial neural network to remove noise from cardiac signals responsively to the received first cardiac signals and the respective noise-added cardiac signals, receiving second cardiac signals captured by at least one second sensing electrode in contact with tissue of a second living subject, and applying the trained artificial neural network to the second cardiac signals to yield noise-reduced cardiac signals.

What is claimed is: 
     
         1 . A method for analyzing signals, comprising:
 receiving first cardiac signals captured by at least one first sensing electrode;   selecting one or more segments of the received first cardiac signals for use as training data;   injecting the one or more selected segments of the received first cardiac signals into a length of wire, which outputs respective noise-added cardiac signals, the noise-added cardiac signals comprising the one or more selected segments of the received first cardiac signals and noise acquired in the length of wire from incident noise of operating electrophysiology equipment;   inputting the noise-added cardiac signals into an artificial neural network;   training the artificial neural network by iteratively comparing the output of the artificial neural network and the one or more segments of the received first cardiac signals and adjusting parameters of the artificial neural network to reduce any difference between an output of the artificial neural network and the received one or more segments of the first cardiac signals;   receiving second cardiac signals captured by at least one second sensing electrode, the second cardiac signals including noise acquired in the length of wire; and   reducing the noise acquired in the length of wire from the second cardiac signals by applying the trained artificial neural network to the second cardiac signals to yield noise-reduced cardiac signals.   
     
     
         2 . The method according to  claim 1 , further comprising converting the first cardiac signals from a digital form to an analog form, the injecting including injecting the first cardiac signals in the analog form into the length of wire, the method further comprising converting the noise-added cardiac signals to the digital form, the training including training the artificial neural network to reduce noise from cardiac signals responsively to the received first cardiac signals in the digital form and the respective noise-added cardiac signals in the digital form. 
     
     
         3 . The method according to  claim 1 , wherein the training comprises training an autoencoder comprising an encoder and a decoder. 
     
     
         4 . The method according to  claim 1 , further comprising rendering to a display a representation of one or more of the noise-reduced cardiac signals. 
     
     
         5 . The method according to  claim 4 , further comprising generating and rendering to the display an electroanatomic map with the one or more of the noise-reduced cardiac signals. 
     
     
         6 . The method according to  claim 1 , further comprising:
 providing a first catheter comprising the at least one first sensing electrode; and   providing a second catheter comprising the at least one second sensing electrode.   
     
     
         7 . The method according to  claim 6 , wherein the first catheter includes the second catheter. 
     
     
         8 . A computer program product, comprising a non-transitory computer-readable medium having computer-readable program code embodied therein to be executed by one or more processors, the program code including instructions to:
 receive first cardiac signals captured by at least one first sensing electrode;   select one or more segments of the received first cardiac signals for use as training data;   inject the one or more selected segments of the received first cardiac signals into a length of wire, which outputs respective noise-added cardiac signals, the noise-added cardiac signals comprising the one or more selected segments of the received first cardiac signals and noise acquired in the length of wire from incident noise of operating electrophysiology equipment;   input the noise-added cardiac signals into an artificial neural network;   train the artificial neural network by iteratively comparing the output of the artificial neural network and the one or more segments of the received first cardiac signals and adjusting parameters of the artificial neural network to reduce any difference between an output of the artificial neural network and the received one or more segments of the first cardiac signals;   receive second cardiac signals captured by at least one second sensing electrode, the second cardiac signals including noise acquired in the length of wire; and   reduce the noise acquired in the length of wire from the second cardiac signals by applying the trained artificial neural network to the second cardiac signals to yield noise-reduced cardiac signals.   
     
     
         9 . The software product according to  claim 8 , wherein the program code includes further instructions to cause a representation of at least one of the noise-reduced cardiac signals to be rendered to a display. 
     
     
         10 . The software product according to  claim 8 , wherein the program code includes further instructions to generate and cause an electroanatomic map to be rendered to a display together with one or more of the noise-reduced cardiac signals. 
     
     
         11 . A medical system, comprising:
 one or more processors; and   a non-transitory computer readable medium storing a plurality of instructions, which when executed, cause the one or more processors to:
 receive first cardiac signals captured by at least one first sensing electrode; 
 select one or more segments of the received first cardiac signals for use as training data; 
 inject the one or more selected segments of the received first cardiac signals into a first end of a length of wire, which outputs at a second end respective noise-added cardiac signals, the noise-added cardiac signals comprising the one or more selected segments of the received first cardiac signals and noise acquired in the length of wire from incident noise of operating electrophysiology equipment; 
 input the noise-added cardiac signals into an artificial neural network; 
 train the artificial neural network by iteratively comparing the output of the artificial neural network and the one or more segments of the received first cardiac signals and adjusting parameters of the artificial neural network to reduce any difference between an output of the artificial neural network and the received one or more segments of the first cardiac signals; 
 receive second cardiac signals captured by at least one second sensing electrode, the second cardiac signals including noise acquired in the length of wire; and 
 reduce the noise acquired in the length of wire from the second cardiac signals by applying the trained artificial neural network to the second cardiac signals to yield noise-reduced cardiac signals. 
   
     
     
         12 . The system according to  claim 11 , wherein the artificial neural network comprises an autoencoder including an encoder and a decoder, the processing circuitry being configured to train the autoencoder to reduce noise from cardiac signals responsively to the received first cardiac signals and the respective noise-added cardiac signals. 
     
     
         13 . The system according to  claim 11 , wherein the processing circuitry further comprises:
 a digital-to-analog converter configured to convert the first cardiac signals from a digital form to an analog form, the processing circuitry being configured to inject the first cardiac signals in the analog form into the length of wire; and   an analog-to-digital converter configured to convert the noise-added cardiac signals to the digital form, the processing circuitry being configured to train the artificial neural network to reduce noise from cardiac signals responsively to the received first cardiac signals in the digital form and the respective noise-added cardiac signals in the digital form.   
     
     
         14 . The system according to  claim 11 , wherein the trained artificial neural network comprises an autoencoder including an encoder and a decoder, the processing circuitry being configured to apply the autoencoder to the second cardiac signals to yield the noise-reduced cardiac signals. 
     
     
         15 . The system according to  claim 11 , further comprising a display, wherein the processing circuitry is configured to render to the display a representation of at least one of the noise-reduced cardiac signals. 
     
     
         16 . The system according to  claim 11 , further comprising a display, wherein the processing circuitry is configured to generate and render to the display an electroanatomic map responsively to the noise-reduced cardiac signals. 
     
     
         17 . The system according to  claim 11 , further comprising:
 a first catheter comprising the at least one first sensing electrode; and   a second catheter comprising the at least one second sensing electrode.   
     
     
         18 . The system according to  claim 17 , wherein the first catheter includes the second catheter.